Dan Hooper - Fermilab/University of Chicago Community Summer Study (Snomass 2013)

On behalf of the CF4 Working Group (Manoj Kaplinghat, Konstantin Matchev, DH) The Dark Matter Discovery Age In a sense, we have already discovered dark matter  Observations of galaxy dynamics, galaxy clusters, large scale structure, the cosmic microwave background, light element abundances, etc. collectively provide an overwhelming body of evidence in favor of the conclusion that most of the matter in our Universe consists of (relatively) cold and collisionless particles  Despite considerable effort, no viable alternatives to this conclusion have been proposed

But we have not yet identified the nature of the particle(s) that make up the dark matter  No particles contained in the standard model are viable dark matter candidates – dark matter is new physics

The Dark Matter Discovery Age In a sense, we have already discovered dark matter  Observations of galaxy dynamics, galaxy clusters, large scale structure, the cosmic microwave background, light element abundances, etc. collectively provide an overwhelming body of evidence in favor of the conclusion that most of the matter in our Universe consists of (relatively) cold and collisionless particles  Despite considerable effort, no viable alternatives to this conclusion have been proposed

After decades of experimental progress, we are currently in a position in which many or most of the best motivated dark matter candidates are within reach of near future experiments – The successful detection of dark matter particles within the next decade seems likely

SLAC 5/25/13 18:39

Cryogenic Solid State CDMS/SuperCDMS EDELWEISS/CRESST/EURECA CoGeNT/C4 TEXONO/CDEX

Liquid Xenon LUX/LZ

XENON PandaX XMASS

Liquid Argon ArDM Darkside DEAP CLEAN

Crystal and Annual Modulation DAMA/LIBRA KIMS ELEGANT ANAIS SLAC 5/25/13 18:39 CINDMS Princeton NaI WIMP Direct Detection CensusDM-Ice Cryogenic Solid State CDMS/SuperCDMS Threshold Detectors EDELWEISS/CRESST/EURECA Technology Description 1. Experiment Status, Target Mass CoGeNT/C4 PICASSO TEXONO/CDEX SIMPLE 2. Fiducial target mass COUPP 3. Backgrounds after passive and Liquid Xenon LUX/LZ Directional Detection

SI Is yourÊ experiment currently operating, and with what total target mass? -43 Ê Crystals (a factor of 2 every 15 months) s 10 Ú ‡ If not, whenÚÊ doÚÊ you expect to operate, and with what total target mass? Threshold Detectors What total targetÚ mass do you expect ÏtoLiquid haveArgon operating 10 years from now? http://www.snowmass2013.org/ tiki- -45 ÚÁ Ì Ú Liquid Xenon Nucleon 10 Ì

Technology Description- 2. Fiducial target mass Ûı Á Threshold Detectors index.php?page =SLAC PICASSO Ì Ù SIMPLE WIMP Û 10-47 Ì COUPP http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5 Û 10-49 1990 2000 2010 2020 Directional Detection Year DRIFT 29 July 2013 NewageCosmic Frontier Intro D.'McKinsey''''''''Direct'Detec,on' 11 DMTPC MIMAC D3

New Ideas DAMIC Liquid helium-4 NEXT Nuclear emulsions (Naka, Japan) DNA & Nano-explosions (Drukier/Cantor)

THE QUESTIONS

1. Experiment Status and Target Mass

Is your experiment currently operating, and with what total target mass? If not, when do you expect to operate, and with what total target mass? What total target mass do you expect to have operating 10 years from now?

2. Fiducial target mass

http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5 SLAC 5/25/13 18:39

Cryogenic Solid State CDMS/SuperCDMS EDELWEISS/CRESST/EURECA CoGeNT/C4 TEXONO/CDEX

Liquid Xenon LUX/LZ

XENON PandaX XMASS

Liquid Argon ArDM Darkside DEAP CLEAN

SI Is yourÊ experiment currently operating, and with what total target mass? -43 Ê Crystals (a factor of 2 every 15 months) s 10 Ú ‡ If not, whenÚÊ doÚÊ you expect to operate, and with what total target mass? Ï Liquid Argon  Threshold Detectors What total targetÚ mass do you expect to have operating 10 years from now? http://www.snowmass2013.org/ Some important benchmarkstiki- exist -45 ÚÁ Ì Ú Liquid Xenon Nucleon 10 Ì

Technology Description- 2. Fiducial target mass Ûı Á Threshold Detectors index.php?pagealong this =SLACline: PICASSO Ì Ù SIMPLE WIMP Û 10-47 Ì  Mid-late 90s: Direct detection COUPP http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5 Û experiments excluded the cross 10-49 sections predicted for a WIMP which 1990 2000 2010 2020 Directional Detection Year scatters and annihilates through DRIFT 29 July 2013 Z-exchange NewageCosmic Frontier Intro D.'McKinsey''''''''Direct'Detec,on' 11 DMTPC  Now!: Current experiments are MIMAC D3 beginning to test WIMPs which interact Higgs-mediated scattering through Higgs exchange New Ideas (A-funnel, focus point neutralinos) (including many SUSY models) DAMIC DM SM DM Liquid helium-4 DM NEXT Nuclear emulsions (Naka, Japan) DNA & Nano-explosions (Drukier/Cantor) DM SM SM SM

THE QUESTIONS

1. Experiment Status and Target Mass Is your experiment currently operating, and with what total target mass? If not, when do you expect to operate, and with what total target mass? What total target mass do you expect to have operating 10 years from now?

2. Fiducial target mass

http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5 SLAC 5/25/13 18:39

Cryogenic Solid State CDMS/SuperCDMS EDELWEISS/CRESST/EURECA CoGeNT/C4 TEXONO/CDEX

Liquid Xenon LUX/LZ

XENON PandaX XMASS

Liquid Argon ArDM Darkside DEAP CLEAN

active shielding. XENON DRIFT PandaX Newage 4. Detector Discrimination XMASS DMTPC MIMAC 5. Energy Threshold D3 Liquid Argon 6. Sensitivity versus WIMP mass ArDM New Ideas Darkside DEAP DAMIC 7. Experimental Challenges CLEAN Liquid helium-4 NEXT 8. Annual Modulation …'and'this'progress'is'expected'to'con,nue.'Nuclear emulsions (Naka, Japan) Direct Detection DNA & Nano-explosions (Drukier/Cantor) Crystal and Annual Modulation 9. Unique Capabilities DAMA/LIBRA KIMS 2  If this rate of progress continues without Evolution of the sSI for a 50 GeV c WIMP -39 THE QUESTIONS 10. Determining WIMP properties ELEGANT 10 the observation of a signal for anotherANAIS ‡ CINDMS ‡ ‡ -41 ~5and years, astrophysical the remaining parameters dark matter Princeton NaI D 10 ‡ ê 2 Ê 1. Experiment Status and Target Mass DM-Ice ÊÊÊ cm Ê Cryogenic Detectors models will be severely constrained @ Ú SI Is yourÊ experiment currently operating, and with what total target mass? -43 Ê Crystals s 10 Ú ‡ If not, whenÚÊ doÚÊ you expect to operate, and with what total target mass?  Ï Liquid Argon In order for WIMPs to evade detectionThreshold Detectors What total targetÚ mass do you expect to have operating 10 years from now? http://www.snowmass2013.org/tiki- -45 ÚÁ Ì Ú Liquid Xenon Nucleon 10 Ì

over this time frame, one must considerTechnology Description - 2. Fiducial target mass Ûı Á Threshold Detectors index.php?page=SLAC PICASSO Ì Ù one or more of the following: SIMPLE WIMP Û 10-47 Ì COUPP http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5  WIMPs which couple almost entirely to Û 10-49 leptons or gauge bosons, rather than to 1990 2000 2010 2020 quarks Directional Detection Year DRIFT  29WIMPs July 2013 which annihilate in the early Newage Cosmic Frontier Intro D.'McKinsey''''''''Direct'Detec,on' 11 DMTPC universe through a highly-tuned MIMAC resonance, or through a highly degenerateD3 It is starting to become co-annihilation difficult to hide our best New Ideas motivated dark matter models  WIMPs which are light enough to fall below DAMIC from direct detection expts. experimental energy thresholds (m <10Liquid GeVhelium-4) X NEXT ~5-10 years from now, very  Non-standard cosmology (ie. low reheating)Nuclear emulsions (Naka, Japan) DNA & Nano-explosions few(Drukier/Cantor WIMP) models will remain out of reach THE QUESTIONS

1. Experiment Status and Target Mass

2. Fiducial target mass

http://www.snowmass2013.org/tiki-index.php?page=SLAC Page 3 of 5 Direct Detection ofAxion 'detec,on:'exis,ng'limits'and'future'projec,ons'Axions  The QCD axion is a natural consequence of the Pecci-Quinn solution to the strong CP problem, and represents a well motivated and viable candidate for dark matter  Axion-photon conversion in the presence of strong magnetic field provides a mechanism to directly search for axion dark matter particles  The mass range predicted for axion

dark matter extends over roughly three D.'McKinsey''''''''Direct'Detec,on' orders of magnitude, from 10-6 to 10-3 eV  ADMX is projected to cover the first of these three decades in its first year of operations, and the second decade over the following two years

Indirect Detection

Some Highlights in Indirect Detection A key benchmark for indirect EGRET, Super Kamiokande searches: HEAT -To be produced with the observed dark matter abundance, a thermal INTEGRAL (511 keV) relic must have an annihilation WMAP Haze (at freeze-out) of σv~3x10-26 cm3/s

-Although many factors could PAMELA enable the dark matter to possess a (Rising e+ fraction) somewhat different cross section IceCube today, many models predict current annihilation rates that are within an Completed order of magnitude or so of this estimate AMS-02 Planck -For the first time, existing experiments are beginning to test WIMPs with an annihilation cross section near this value Status of Indirect Detection 18 ACT DM Constraints Gamma-Rays γγ  The Fermi Gamma-Ray Space Telescope’s observations of dwarfHESS VERITAS spheroidal galaxies and the Galactic Center are sensitive to simple thermal WIMPs lighter than ~30 GeV (weaker but competitive limits have Hooper et al. 2012, been derived from clusters, the Ackermann et al. 2011 isotropic gamma-ray background, and 5 subhalo searches) VERITAS DEEP OBSERVATIONS OF THE DWARF ... PHYSICAL REVIEW D 85, 062001 (2012) -3 GC Limits Segue Dwarf Galaxy Limits 2!10 -22 ]

-1 10

Bg Source region FIG. 14: A comparison of the upper limits on the dark matter annihilation cross section derived in this work to those from s 3 ,F  Ground based telescopes other gamma-ray observations. In particular, we show the constraints derived from the observations of dwarf spheroidal -23 Src galaxies [1, 2], the isotropic gamma-ray background [8], and from the Fornax galaxy cluster [3]. If we adopt an NFW halo Background region 10 *F (HESS, VERITAS, MAGIC) profile (or an Einasto or contracted profile), the constraints derived from the Galactic Center are the most stringent. Only if v > [cm 2.7 $ the dark matter halo profile of the Milky Way has a significant core (while dwarf galaxies retain their cusps) are constraints -3 -24 E 10 < from10 dwarfs more stringent. The constraints from the Galactic Center are, for all dark matter masses, more stringent than 9!10-4 are most sensitive to high -4 those reliably extracted from the isotropic gamma-ray background or from galaxy clusters. 8 10 -25 ! 10 7!10-4 mass WIMPs 6!10-4 the results presented here are in no way in conflict with multi-wavelength studies of the Galactic Center, and 10-26 5!10-4 those presented previously which find that annihilating progress from hydrodynamical simulations of dark mat- dark matter can provide a good fit to the observed emis- ter in Milky Way-like galaxies, could further strengthen -0.5 0 0.5 1log (E[TeV])1.5 -27 sion10 [11–13, 16]. In particular, Fermi’s GalacticEinasto Cen-(this work) the dark matter constraints that can be derived from the NFW (this work) 4 ter observations, coupled with observations of the Milky inner Galaxy.

res Sgr Dwarf -28 F 3 Way’s10 radio ﬁlaments, are most easily explainedWillman by 1 a " Ursa Minor / 2 dark matter particle with a mass of mDM Draco7 10

res ⇡26 3 1 -29 F GeV,10 anHESS annihilation Collaboration cross section of v 10 cm /s 0 ⇠ to charged leptons,-1 and distributed in a somewhat con- 10 1.3 1 10 -1 tracted profile (⇢ r ). Acknowledgements: We would like to thank Tim Lin- m# [TeV] -2 Looking toward/ the future, we find very promising den, Mariangela Lisanti, and Keith Bechtol for insightful -3 (Aharonianthe possibility et al. of thefor the post-Fermi HESS gamma-raycollaboration, satellite, PRL 106,comments 1301) as well as Gianfranco(Aliu et al. Bertone, for the Miguel VERITAS Pato, collaboration, PRD 85, 062001) -4 FIG.GAMMA-400 4. Upper [59]. limits As GAMMA-400’s (at 95% CL) overall on the e↵ective velocity-weightedFabio Iocco andFIG. Philippe 5 (color Jetzer online). for providing 95% the CL con- exclusion curves from the VERITAS observations of Segue 1 on v =S as a function of the dark -0.5 0 0.5 1log (E[TeV]) 1.5 h i annihilationarea andSnowmass acceptance cross-section will2013 be comparable " σ v # as ato function that of Fermi, of the tours DM usedCF2: par- in Indirectmatter our Fig. particle 2.Detection DH ismass, supported in the by framework the US of two James models Buckley with a Sommerfeld enhancement. The expected Sommerfeld enhancement S it will likely not be more sensitive to dark matter anni- Department ofapplied Energy. to CK the is particular supported case by aof Fermi- Segue 1 has been computed assuming a Maxwellian dark matter relative velocity distribution. The grey ticlehilations mass frommχ flux-limitedfor the sources, Einasto such and as dwarf NFW galax- densitylab profiles. Fellowship in Theoretical Physics. FSQ is supported FIG. 3. Top panel: Reconstructed differential flux FSrc/Bg, The best1 sensitivity is achieved at mχ ∼ 1 TeV. For com- band area represents a range of generic values for the annihilation cross-section in the case of thermally produced dark matter. Left: 2.7 ies. With considerable improvements in both angular and by Coordenacao de Aperfeisoamento de Pessoal de Nivel weighted with E for better visibility, obtained for the source parison,energy resolution the best relative limits to derived Fermi, however, from observations GAMMA- Superior of dwarf (CAPES).model This I with work winolike was supported neutralino in part dark by matter annihilating to a pair of WþWÀ bosons. Right: model II with a 250 MeV scalar particle and background regions as defined in the text. The units are galaxies400 should at be very able to high much energies, better separate i.e. astrophysical Sgr Dwarf [10],the National Will- Sciencedecaying Foundation into either undere Grante or No. PHY- . See text for further details. 1.7 −2 −1 −1 þ À þ À TeV m s sr .Duetoanenergy-dependentselection manbackgrounds 1, Ursa in Minor the inner [15] Galaxy and from Draco any [9], dark using mat- in1066293. all cases We thank the Aspen Center for Physics for ter annihilation signal that is present. Furthermore, their hospitality. efficiency and the use of effective areas obtained from γ-ray NFW shaped DM profiles, are shown. Similar to source re- simulations, the reconstructed spectra are modified compared gion of the current analysis, dwarf galaxies are objects free circumvent the helicity suppression of the annihilation located at different dark matter particle masses and with to the cosmic-ray power-law spectrum measured on Earth. of astrophysical background sources. The green points rep- cross-section into light leptons, the neutralino can oscillate different amplitudes with respect to model I, because the Bottom panel: Flux residua Fres/∆Fres,whereFres = FSrc − resent[1] A. DarkSUSY Geringer-Sameth, models S. M. Koushiappas [32], which Phys. are Rev. in Lett. agreement107 with, 241302with (2011) charginos [arXiv:1108.3546 Æ, which [astro-ph.HE]]; themselves can preferentially coupling and mass of the exchanged particle differ. Two FBg and ∆Fres is the statistical error on Fres.Theresidual WMAP107, and 241303 collider (2011) [arXiv:1108.2914 constraints [astro-ph.CO]]. and were obtained withC. Farnier a annihilateet al. [Fermi-LAT into Collaboration], leptons. The Nucl. transition In- to a chargino state is channels in which the scalar particle decays either to eþeÀ flux is compatible with a null measurement. Comparable null random[2] The scan Fermi-LAT of the Collaboration, mSUGRA Phys. parameter Rev. Lett. space usingstrum. the Meth. A 630,143(2011). 0 residuals are obtained when varying the radius of the source mediated by the exchange of a Z boson (mZ0 90 GeV, or þÀ have been considered. VERITAS observations following parameter ranges: 10 GeV particle physics and/or astro- Science and Higher Education, the South African De- ! F regulates the dark matter density before freezeout [98]. physical boost) as a function of the dark matter particle partment of Science and Technology and National Re- Figure 5 shows the VERITAS constraints for each of mass is presented. The constraints are then compared to the AsearchforaVHEγ-ray signal from DM annihilations search Foundation, and by the University of Namibia. was conducted using H.E.S.S. data from the GC region. these models, derived with the observations of Segue 1. recent cosmic ray lepton data. We appreciate the excellent work of the technical support The dashed curves show the 95% CL exclusion limits A circular region of radius 1◦ centered at the GC was cho- Following [99], we assume that dark matter annihilates staff in Berlin, Durham, Hamburg, Heidelberg, Palaiseau, without the Sommerfeld correction to the annihilation sen for the search, and contamination by astrophysical exclusively into muons with an annihilation cross-section Paris, Saclay, and in Namibia in the construction and op- cross-section, whereas the solid curves are the limits 26 3 1 γ-ray sources along the Galactic plane was excluded. An v 3 10À cm sÀ . In such a case, we use the eration of the equipment. to the Sommerfeld enhanced annihilation cross-section. h i ¼ Â optimized background subtraction technique was devel- dashed exclusion curve of Fig. 3 (right) to compute The left panel of Fig. 5 shows the constraints on model I, oped and applied to extract the γ-ray spectrum from the 95% CL limits on BF. Figure 6 shows the 95% CL ULs for the annihilation of neutralinos into W W through the source region. The analysis resulted in the determination þ À on the overall boost factor BF. The blue and red shaded exchange of a Z0 boson. The Sommerfeld enhancement of stringent upper limits on the velocity-weighted DM an- regions are the 95% CL contours that best fit the Fermi- exhibits two resonances in the considered dark matter nihilation cross-section !σv", being among the best so far ∗ supported by CAPES Foundation, Ministry of Education LAT and PAMELA eþeÀ data, respectively. The grey particle mass range, for m 4:5 TeV and m shaded area shows the 95% CL excluded region derived at very high energies. At the same time, the limits do not of Brazil ’ ’ † 17 TeV, respectively. VERITAS excludes these reso- differ strongly between NFW and Einasto parametriza- [email protected] from the H.E.S.S. eþeÀ data [99]. The black dot is an ‡ supported by Erasmus Mundus, External Cooperation nances, which boost the annihilation cross-section far be- example of a model which simultaneously fits well the tions of the DM density profile of the Milky-Way. 26 3 1 Window yond the canonical v 3 10À cm sÀ . The right H.E.S.S., PAMELA and Fermi-LAT data. The VERITAS The support of the Namibian authorities and of the § h i Â [email protected] panel of Fig. 5 shows the VERITAS constraints on model VHE -ray observations of Segue 1 rule out a significant University of Namibia in facilitating the construction and ¶ UMR 7164 (CNRS, UniversitéParis VII, CEA, Observa- II, for a scalar particle with mass m 250 MeV. The portion of the regions preferred by cosmic ray lepton data. operation of H.E.S.S. is gratefully acknowledged, as is the toire de Paris) ¼ Sommerfeld enhancement exhibits many more resonances, However, the electron and positron constraints depend on

062001-9 Status of Indirect Detection

Cosmic Rays  AMS and PAMELA are sensitive to antiprotons and positrons from dark matter annihilations in the halo; in some cases competitive with gamma-ray constraints

Neutrinos  IceCube is sensitive to dark matter annihilations in the core of the Sun; constraints on spin-dependent scattering are competitive with those from direct detection experiments

Other Techniques  Radio and X-ray telescopes, observations of the CMB, and other probes can also constrain the dark matter’s annihilation cross section

Dark MatterLHC atLHC the LHC WIMP WIMP Production Production Two basic dark matter strategies: LHC can produce WIMP siblings, LHC can produce WIMP siblings, 1) Pair produce strongly interacting particles χ which decay into WIMPs and which decay into dark matter particles, along χ whichother decay SM intoparticles. WIMPs and with standard model particles (ie. squark/gluino other SM particles. production, followed by their decay to the LSP) “KK Sgluquarkino Pair Production Followed“KK Sgluquarkino by Decay into Pair WIMPs” Production

χ Followed by Decay into WIMPs” χ 2 msibling 2 m sibling χ 2) Produce dark matter pairs directly, along χ with a single jet or photon } LHC can directly produce WIMP pairs. } LHC can directly produce WIMP pairs. χ χ

2 mχ 2 mχ } No WIMP Production. No WIMP Production. (partonic) } Energy(partonic) Energy }} Dark MatterLHC atLHC the LHC WIMP WIMP Production Production Two basic dark matter strategies: LHC can produce WIMP siblings, LHC can produce WIMP siblings, 1) Pair produce strongly interacting particles χ which decay into WIMPs and which decay into dark matter particles, along χ whichother decay SM intoparticles. WIMPs and with standard model particles (ie. squark/gluino other SM particles. production, followed by their decay to the LSP) “KK Sgluquarkino Pair Production Followed“KK Sgluquarkino by Decay into Pair WIMPs” Production -Potentially high rates χ Followed by Decay into WIMPs” -Prospects depend on masses of the new χ strongly interacting particles (model2 msibling dependent) 2 msibling χ 2) Produce dark matter pairs directly, along χ with a single jet or photon } LHC can directly produce WIMP pairs. -Lower rates (limited mass reach) } LHC can directly produce WIMP pairs. χ -Less model dependent χ

2 mχ 2 mχ } No WIMP Production. No WIMP Production. (partonic) } Energy(partonic) Energy }} Why So Many Techniques?

 If we knew what the dark matter consisted of, we could optimize our strategy to observe it

 But we don’t. And different experimental approaches are sensitive to dark matter candidates with different characteristics, and provide us with different

types of information – complementarity! 3

Dark Matter

Leptons Nuclear Matter Photons, Other dark electrons, muons, quarks, gluons W, Z, h bosons particles taus, neutrinos

DM DM DM SM SM DM DM DM Direct Indirect Particle Astrophysical Detection Detection Colliders Probes SM SM DM SM SM DM DM DM

FIG. 1: Dark matter may have non-gravitational interactions with one or more of four categories of particles: nuclear matter, leptons, photons and other bosons, and other dark particles. These interactions may then be probed by four complementary approaches: direct detection, indirect detection, particle colliders, and astrophysical probes. The lines connect the experimental approaches with the categories of particles that they most stringently probe (additional lines can be drawn in speciﬁc model scenarios). The diagrams give example reactions of dark matter with standard model particles (SM) for each experimental approach.

axions through their interaction with photons in a magnetic field. Indirect Detection. Pairs of dark matter particles annihilate producing high-energy particles • (antimatter, neutrinos, or photons). Alternatively, dark matter may be metastable, and its decay may produce the same high-energy particles. Particle Colliders. Particle colliders, such as the Large Hadron Collider (LHC) and proposed • future colliders, produce dark matter particles, which escape the detector, but are discovered as an excess of events with missing energy or momentum. Astrophysical Probes. The particle properties of dark matter are constrained through its • impact on astrophysical observables. In particular, dark matter’s non-gravitational interactions could observably impact the densities of dark matter present in the central regions of galaxies, or the amount of dark matter substructure found in halos. Such interactions may also alter the cooling rates of stars, and influence the pattern of temperature fluctuations observed in the cosmic microwave background.

These search strategies are each shown in Fig. 1 and are connected to the particle interactions that they most stringently probe. After discussing in more detail many of the most promising particle candidates for dark matter in Sec. II, we return in Sec. III to these four pillars in more detail, discussing their current status and future prospects. In Sec....

II. DARK MATTER CANDIDATES

In this section, we briefly summarize a number of specific dark matter candidates and candidate classes that have been considered in the literature. While certainly not exhaustive, this discussion is intended to reflect a representative sample of how the particle physics community currently views the form that dark matter particles might take. Why So Many Techniques? No one technique will answer all (or even most) of our questions about the nature of dark matter:

 The LHC cannot tell us whether a weakly interacting particle is stable and cosmologically relevant, or merely long-lived on collider timescales

 Neither direct not indirect detection experiments alone will be able to determine the spin of the dark matter candidate, or identify how these particles interact (although they can narrow down the possibilities)

 These techniques each suffer from different uncertainties and limitations

 By combining information from multiple techniques, a much more detailed description of the dark matter could emerge

R-parity NMSSM MSSM violating

WIMPless DM Supersymmetry Hidden Sector DM Gravitino DM mSUGRA Self-Interacting DM pMSSM

Q-balls Techni- baryons Dark Photon R-parity Conserving Dirac Asymmetric DM DM

Light Extra Dimensions Force Carriers

Warm DM Soliton DM Theories of UED DM Sterile Neutrinos Quark 6d Warped Extra Nuggets Dimensions Dark Matter 5d

RS DM T-odd DM Axion DM ?

Little Higgs QCD Axions

Axion-like Particles T Tait Littlest Higgs 12

Indirect Detection is sensitive to dark matter interactions with all standard model particles, • directly probes the annihilation process suggested by the WIMP miracle, and experimental sensitivities are expected to improve greatly on several fronts in the coming decade. Discovery through indirect detection requires understanding astrophysical backgrounds and the signal strength is typically subject to uncertainties in halo proﬁles. Indirect detection signals are absent if dark matter annihilation is insigniﬁcant now, for example, as in the case of asymmetric dark matter. Particle Colliders provide the opportunity to study dark matter in a highly controlled labora- • tory environment, may be used to precisely constrain many dark matter particle properties, and are sensitive to the broad range of masses favored for WIMPs. Hadron colliders are relatively insensitive to dark matter that interacts only with leptons, and colliders are unable to distinguish missing momentum signals produced by a particle with lifetime 100 ns from ⇠ one with lifetime > 1017 s, as required for dark matter. Astrophysical Probes⇠ are unique probes of the “warmth” of dark matter and hidden dark • matter properties, such as its self-interaction strength, and they measure the e↵ects of dark matter properties on structure formation in the Universe. Astrophysical probes are typi- Complementaritycally unable to distinguish various forms of CDM of from Direct, each other or make Indirect, other precision measurements of the particle properties of dark matter. 13

and ColliderV. QUANTITATIVE Searches COMPLEMENTARITY for Dark Matter

A. E↵ective Operator Description

The qualitative features outlined above may be illustrated in a simple and fairly model- independent setting by considering dark matter that interacts with standard model particles through four-particle contact interactions, which represent the exchange of very heavy particles. These contact interactions are expected to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest. To do this, we may choose representative, generation-independent, couplings of a spin-1/2 dark matter particle with quarks q, gluons g, and leptons ` (including neutrinos) given by 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) M 2 5 µ 5 M 3 µ⌫ M 2 µ q q g ` X X` The interactions with quarks mediate spin-dependent direct signals, whereas those with gluons mediate spin-independent direct signals. The coecients M , M , and M characterize the strength FIG. 2: Dark matter discovery prospects in the (m, /th)q planeg for current` and future direct detection [98], indirectof the detection interaction [99, with 100], the respective and particle standard colliders model [101–103] particle, and for in dark this representative matter coupling example to gluons [104], quarks should [104, be 105], chosen and such leptons that the [106, combined 107], as annihilation indicated. cross section into all three channels provides the correct relic density of dark matter. The values of the three interaction strengths, together with the mass of the dark matter particle m, completely deﬁne this theory and allow one to predict the rate of both spin-dependent and spin-independent direct scattering, the annihilation cross section above th (green-shaded regions in Fig. 2), it would have discovered one species of dark matter, which, into however, quarks, gluons, could and not leptons, account and forthe production all of the rate dark of matter dark matter (within at colliders. this model framework), and Each class of dark matter search outlined in Sec. III is sensitive to some range of the interaction consequentlystrengths for point a given to dark other matter dark mass. matter Therefore, species they still are waiting all implicitly to be putting discovered. a bound on the Inannihilation Fig. 2, we cross assemble section into the a discovery particular channel. potential Since and the current annihilation bounds cross for section several predicts near-term dark matterthe dark searches matter that relic are density, sensitive the reach to interactionsof any experiment with is quarks thus equivalent and gluons, to a fraction or leptons. of the It is clear thatobserved the searches dark matter are complementary density. This connection to each can other be seen in terms in the plots of being in Fig. sensitive 2, which toshow interactions the with di↵erent standard model particles. These results also illustrate that within a given interaction type, the reach of di↵erent search strategies depends sensitively on the dark matter mass. For example, direct searches for dark matter are very powerful for masses around 100 GeV, but have diculty at very low masses, where the dark matter particles carry too little momentum to noticeably a↵ect heavy nuclei. This region of low mass is precisely where collider production of dark matter is easiest, since high energy collisions readily produce light dark matter particles with large momenta.

B. Complete Models

1. pMSSM (SLAC)

and ColliderV. QUANTITATIVE Searches COMPLEMENTARITY for Dark Matter

A. E↵ective Operator Description

The qualitative features outlined above may be illustrated in a simple and fairly model- independent setting by considering dark matter that interacts with standard model particles through four-particle contact interactions, which represent the exchange of very heavy particles. These contact interactions are expected to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest. To do this, we may choose representative, generation-independent, couplings of a spin-1/2 dark matter particle with quarks q, gluons g, and leptons ` (including neutrinos) given by 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) M 2 5 µ 5 M 3 µ⌫ M 2 µ q q g ` X X` The interactions with quarks mediate spin-dependent direct signals, whereas those with gluons mediate spin-independent direct signals. The coecients M , M , and M characterize the strength FIG. 2:In Darkthis mattercase, discoverythe WIMP prospects will be in thediscovered (m, /th )byq plane bothg for future current` direct and future experiments direct detection and [98], indirectofby the detectionthe interaction LHC, [99, withbut 100], thenot respective and by particleindirect standard colliders detection model [101–103] particle, efforts and for in dark this representative matter coupling example to gluons [104], quarksshould [104, be 105], chosen and such leptons that the [106, combined 107], as annihilation indicated. cross section into all three channels provides  theThese correct signals relic density strong of dark favor matter. a The thermal values of relic the three interpretation, interaction strengths, without together significant with theannihilations mass of the dark through matter particle otherm couplings, completely deﬁne this theory and allow one to predict the rate of both spin-dependent and spin-independent direct scattering, the annihilation cross section above th (green-shaded regions in Fig. 2), it would have discovered one species of dark matter, which, into however, quarks, gluons, could and not leptons, account and forthe production all of the rate dark of matter dark matter (within at colliders. this model framework), and Each class of dark matter search outlined in Sec. III is sensitive to some range of the interaction consequently strengths for point a given to dark other matter dark mass. matter Therefore, species they still are waiting all implicitly to be putting discovered. a bound on the Inannihilation Fig. 2, we cross assemble section into the a discovery particular channel. potential Since and the current annihilation bounds cross for section several predicts near-term dark matter the dark searches matter that relic are density, sensitive the reach to interactionsof any experiment with is quarks thus equivalent and gluons, to a fraction or leptons. of the It is clear that observed the searches dark matter are complementary density. This connection to each can other be seen in terms in the plots of being in Fig. sensitive 2, which toshow interactions the with di↵erent standard model particles. These results also illustrate that within a given interaction type, the reach of di↵erent search strategies depends sensitively on the dark matter mass. For example, direct searches for dark matter are very powerful for masses around 100 GeV, but have diculty at very low masses, where the dark matter particles carry too little momentum to noticeably a↵ect heavy nuclei. This region of low mass is precisely where collider production of dark matter is easiest, since high energy collisions readily produce light dark matter particles with large momenta.

B. Complete Models

1. pMSSM (SLAC)

and ColliderV. QUANTITATIVE Searches COMPLEMENTARITY for Dark Matter

A. E↵ective Operator Description

The qualitative features outlined above may be illustrated in a simple and fairly model- independent setting by considering dark matter that interacts with standard model particles through four-particle contact interactions, which represent the exchange of very heavy particles. These contact interactions are expected to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest. To do this, we may choose representative, generation-independent, couplings of a spin-1/2 dark matter particle with quarks q, gluons g, and leptons ` (including neutrinos) given by 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) M 2 5 µ 5 M 3 µ⌫ M 2 µ q q g ` X X` The interactions with quarks mediate spin-dependent direct signals, whereas those with gluons mediate spin-independent direct signals. The coecients M , M , and M characterize the strength FIG. 2:In Darkthis mattercase, discoverythe WIMP prospects will be in thediscovered (m, /th )byq plane bothg for future current` direct and future experiments direct detection and [98], indirectofby the detectionthe interaction LHC, [99, withbut 100], thenot respective and by particleindirect standard colliders detection model [101–103] particle, efforts and for in dark this representative matter coupling example to gluons [104], quarksshould [104, be 105], chosen and such leptons that the [106, combined 107], as annihilation indicated. cross section into all three channels provides  theThese correct signals relic density strong of dark favor matter. a The thermal values of relic the three interpretation, interaction strengths, without together significant with theannihilations mass of the dark through matter particle otherm couplings, completely deﬁne this theory and allow one to predict the rate of both spin-dependent and spin-independent direct scattering, the annihilation cross section above th (green-shaded regions in Fig. 2), it would have discovered one species of dark matter, which, intoIn however, quarks,other gluons,cases, could and a not leptons,discovery account and forthe could production all of be the made, rate dark of matter darkbut matterless (within supporting at colliders. this model of the framework), thermal and relicEach interpretation class of dark matter – searchOther outlined important in Sec. IIIcouplings? is sensitive to Non-thermal some range of the origin? interaction consequentlystrengths for point a given to dark other matter dark mass. matter Therefore, species they still are waiting all implicitly to be putting discovered. a bound on the Inannihilation Fig. 2, we cross assemble section into the a discovery particular channel. potential Since and the current annihilation bounds cross for section several predicts near-term dark matterthe dark searches matter that relic are density, sensitive the reach to interactionsof any experiment with is quarks thus equivalent and gluons, to a fraction or leptons. of the It is clear that observed the searches dark matter are complementary density. This connection to each can other be seen in terms in the plots of being in Fig. sensitive 2, which toshow interactions the with di ↵erent standard model particles. These results also illustrate that within a given interaction type, the reach of di↵erent search strategies depends sensitively on the dark matter mass. For example, direct searches for dark matter are very powerful for masses around 100 GeV, but have diculty at very low masses, where the dark matter particles carry too little momentum to noticeably a↵ect heavy nuclei. This region of low mass is precisely where collider production of dark matter is easiest, since high energy collisions readily produce light dark matter particles with large momenta.

B. Complete Models

1. pMSSM (SLAC)

ComplementarityIndirect of Detection Direct,is sensitive to dark matterIndirect, interactions with all standard model particles, • directly probes the annihilation process suggested by the WIMP miracle, and experimental sensitivities are expected to improve greatly on several fronts in the coming13 decade. Discovery and Collider Searchesthrough indirect detectionfor requires Dark understanding Matter astrophysical backgrounds and the signal strength is typically subject to uncertainties in halo proﬁles. Indirect detection signals are absent if dark matter annihilation is insigniﬁcant now, for example, as in the case of asymmetric dark matter. Particle Colliders provide the opportunity to study dark matter in a highly controlled labora- • tory environment, may be used to precisely constrain many dark matter particle properties, and are sensitive to the broad range of masses favored for WIMPs. Hadron colliders are relatively insensitive to dark matter that interacts only with leptons, and colliders are unable to distinguish missing momentum signals produced by a particle with lifetime 100 ns from ⇠ one with lifetime > 1017 s, as required for dark matter. Astrophysical Probes⇠ are unique probes of the “warmth” of dark matter and hidden dark • matter properties, such as its self-interaction strength, and they measure the e↵ects of dark matter properties on structure formation in the Universe. Astrophysical probes are typically unable to distinguish various forms of CDM from each other or make other precision measurements of the particle properties of dark matter.

V. QUANTITATIVE COMPLEMENTARITY

A. E↵ective Operator Description FIG. 2: Dark matter discovery prospects in the (m , / ) plane for current and future direct detection [98],  Alternatively, signals could be seenThe in qualitative a thcombinations features outlined of above indirect may be illustratedexperiments in a simple and fairly model- indirectwith detection either [99,direct 100], experiments and particle collidersorindependent the LHC, [101–103] setting again by considering for either dark dark matterwith matter or coupling thatwithout interacts to gluons with standard [104], model particles quarks [104, 105], and leptons [106, 107], asthrough indicated. four-particle contact interactions, which represent the exchange of very heavy particles. consistency with a simple thermalThese origin contact interactions are expected to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest. To do this, we may choose representative, generation-independent, couplings of a spin-1/2 dark above th (green-shaded regions in Fig.matter 2), it particle would with have quarks discoveredq, gluons g, andone leptons species` (including of dark neutrinos) matter, given by which, however, could not account for all of the dark matter (within this model framework), and 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) consequently point to other dark matter species stillM waiting2 5 to beµ 5 discovered.M 3 µ⌫ M 2 µ q q g ` ` In Fig. 2, we assemble the discovery potential and currentX bounds for several near-termX dark matter searches that are sensitive to interactionsThe interactions with with quarks quarks mediate and spin-dependentgluons, or leptons. direct signals, It whereas is clear those with gluons mediate spin-independent direct signals. The coecients Mq, Mg, and M` characterize the strength that the searches are complementary to eachof the interaction other in with terms the respective of being standard sensitive model particle,to interactions and in this representative with example di↵erent standard model particles. Theseshould results be chosen also such illustrate that the combined that within annihilation a given cross interaction section into all three type, channels provides the correct relic density of dark matter. The values of the three interaction strengths, together with the reach of di↵erent search strategies dependsthe mass of sensitively the dark matter on particle them dark, completely matter deﬁne mass. this theory For example, and allow one to predict the direct searches for dark matter are veryrate powerful of both spin-dependent for masses and around spin-independent 100 GeV, direct but scattering, have the di annihilationculty cross section into quarks, gluons, and leptons, and the production rate of dark matter at colliders. at very low masses, where the dark matterEach particles class of darkcarry matter too search little outlined momentum in Sec. III is to sensitive noticeably to some rangea↵ect of the interaction heavy nuclei. This region of low mass is preciselystrengths for where a given collider dark matter production mass. Therefore, of dark they are matter all implicitly is easiest, putting a bound on the annihilation cross section into a particular channel. Since the annihilation cross section predicts since high energy collisions readily producethe dark light matter dark relic matter density, the particles reach of any with experiment large momenta.is thus equivalent to a fraction of the observed dark matter density. This connection can be seen in the plots in Fig. 2, which show the

B. Complete Models

1. pMSSM (SLAC)

V. QUANTITATIVE COMPLEMENTARITY

 In all of the cases shown, at least one technique will discoverA. E↵ective the OperatorWIMP Descriptionif lighter than FIG. 2: Dark matter discovery prospects in the (m , / ) plane for current and future direct detection [98], ~1 TeV (exceptions to this conclusionThe exist, qualitative butth they features are outlined exceptions) above may be illustrated in a simple and fairly model- indirect detection [99, 100], and particle collidersindependent [101–103] setting by considering for dark dark matter matter coupling that interacts to gluons with standard [104], model particles  quarksThe [104, relative 105], and discovery leptons [106,prospects 107], as forthrough indicated. direct, four-particle indirect, contact and interactions, LHC searches which represent are very the exchange model of very heavy particles. dependent – any could get there first,These and contact any interactions could see are expected nothing to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest. To do this, we may choose representative, generation-independent, couplings of a spin-1/2 dark above th (green-shaded regions in Fig.matter 2), it particle would with have quarks discoveredq, gluons g, andone leptons species` (including of dark neutrinos) matter, given by which, however, could not account for all of the dark matter (within this model framework), and 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) consequently point to other dark matter species stillM waiting2 5 to beµ 5 discovered.M 3 µ⌫ M 2 µ q q g ` ` In Fig. 2, we assemble the discovery potential and currentX bounds for several near-termX dark matter searches that are sensitive to interactionsThe interactions with with quarks quarks mediate and spin-dependentgluons, or leptons. direct signals, It whereas is clear those with gluons mediate spin-independent direct signals. The coecients Mq, Mg, and M` characterize the strength that the searches are complementary to eachof the interaction other in with terms the respective of being standard sensitive model particle,to interactions and in this representative with example di↵erent standard model particles. Theseshould results be chosen also such illustrate that the combined that within annihilation a given cross interaction section into all three type, channels provides the correct relic density of dark matter. The values of the three interaction strengths, together with the reach of di↵erent search strategies dependsthe mass of sensitively the dark matter on particle them dark, completely matter deﬁne mass. this theory For example, and allow one to predict the direct searches for dark matter are veryrate powerful of both spin-dependent for masses and around spin-independent 100 GeV, direct but scattering, have the di annihilationculty cross section into quarks, gluons, and leptons, and the production rate of dark matter at colliders. at very low masses, where the dark matterEach particles class of darkcarry matter too search little outlined momentum in Sec. III is to sensitive noticeably to some rangea↵ect of the interaction heavy nuclei. This region of low mass is preciselystrengths for where a given collider dark matter production mass. Therefore, of dark they are matter all implicitly is easiest, putting a bound on the annihilation cross section into a particular channel. Since the annihilation cross section predicts since high energy collisions readily producethe dark light matter dark relic matter density, the particles reach of any with experiment large momenta.is thus equivalent to a fraction of the observed dark matter density. This connection can be seen in the plots in Fig. 2, which show the

B. Complete Models

1. pMSSM (SLAC)

V. QUANTITATIVE COMPLEMENTARITY

A. E↵ective Operator Description  In all of the cases shown, at least one technique will discover the WIMP if lighter than FIG. 2: Dark matter discovery prospects in the (m , / ) plane for current and future direct detection [98], ~1 TeV (exceptions to this conclusionThe exist, qualitative butth they features are outlined exceptions) above may be illustrated in a simple and fairly model- indirect detection [99, 100], and particle collidersindependent [101–103] setting by considering for dark dark matter matter coupling that interacts to gluons with standard [104], model particles  quarksThe [104, relative 105], and discovery leptons [106,prospects 107], as forthrough indicated. direct, four-particle indirect, contact and interactions, LHC searches which represent are very the exchange model of very heavy particles. dependent – any could get there first,These and contact any interactions could see are expected nothing to work well to describe theories in which the exchanged particle mass is considerably larger than the momentum transfer of the physical process of interest.  We will likely need multiple types of observationsTo do this, we may to choose have representative, any chance generation-independent, of identifying couplings the of a spin-1/2 dark aboveWIMPth (green-shaded – for example, regions the three in Fig. pointsmatter 2), shown it particle would lookwith have the quarks discovered sameq, gluons tog direct, andone leptons species detection,` (including of dark but neutrinos) have matter, given by which, however, could not account for all of the dark matter (within this model framework), and very different indirect and collider signals 1 ↵ 1 ¯ µ q¯ q + S ¯ Gaµ⌫Ga + ¯ µ `¯ ` . (1) consequently point to other dark matter species stillM waiting2 5 to beµ 5 discovered.M 3 µ⌫ M 2 µ q q g ` ` In Fig. 2, we assemble the discovery potential and currentX bounds for several near-termX dark matter searches that are sensitive to interactionsThe interactions with with quarks quarks mediate and spin-dependentgluons, or leptons. direct signals, It whereas is clear those with gluons mediate spin-independent direct signals. The coecients Mq, Mg, and M` characterize the strength that the searches are complementary to eachof the interaction other in with terms the respective of being standard sensitive model particle,to interactions and in this representative with example di ↵erent standard model particles. Theseshould results be chosen also such illustrate that the combined that within annihilation a given cross interaction section into all three type, channels provides the correct relic density of dark matter. The values of the three interaction strengths, together with the reach of di↵erent search strategies dependsthe mass of sensitively the dark matter on particle them dark, completely matter deﬁne mass. this theory For example, and allow one to predict the direct searches for dark matter are veryrate powerful of both spin-dependent for masses and around spin-independent 100 GeV, direct but scattering, have the di annihilationculty cross section into quarks, gluons, and leptons, and the production rate of dark matter at colliders. at very low masses, where the dark matterEach particles class of darkcarry matter too search little outlined momentum in Sec. III is to sensitive noticeably to some rangea↵ect of the interaction heavy nuclei. This region of low mass is preciselystrengths for where a given collider dark matter production mass. Therefore, of dark they are matter all implicitly is easiest, putting a bound on the annihilation cross section into a particular channel. Since the annihilation cross section predicts since high energy collisions readily producethe dark light matter dark relic matter density, the particles reach of any with experiment large momenta.is thus equivalent to a fraction of the observed dark matter density. This connection can be seen in the plots in Fig. 2, which show the

B. Complete Models

1. pMSSM (SLAC)

The e↵ective theory description (1) of the dark matter interactions with standard model particles is an attempt to capture the salient features of the dark matter phenomenology without reference to any specific theoretical model. However, the complementarity between the di↵erent dark matter probes seen in Fig. 2 persists also when one considers specific well-motivated theoretical models. Among the many possible alternatives, low energy supersymmetry [108] has been the most popular and widely studied extension of the standard model, and we shall use it here as our second example. 0 In supersymmetry, the DM candidate is generally the lightest neutralino ˜1, which is its own anti- particle. Even within the general framework of supersymmetry, there are many di↵erent model scenarios, distinguished by a number of input parameters ( 20). A model-independent approach to ⇠ supersymmetry is to scan over all those input parameters and consider all models that pass all existing experimental constraints and have a dark matter candidate which could explain at least a portion of the observed dark matter density [109]. Results from such model-independent scans with over 200,000 points are shown in Fig. 3, where each dot represents one particular supersym- metric model. Within each model, the dark matter interactions are completely specified and one Summary  Direct and indirect astrophysical searches for dark matter, as well searches at the LHC, are each at or are approaching the sensitivity expected to be required to detect dark matter (in the form of WIMPs) non-gravitationally for the first time  Similar prospects exist for dark matter in the form of axions  If ~5-10 years pass without discovery, we will have to radically revise our ideas about the nature of dark matter  Even if they succeed in detecting dark matter, none of these techniques alone is likely to tell us much about the nature of these particles  These techniques each offer different information, have different uncertainties, and are sensitive to different types of dark matter candidates  By taking advantage of the complementarity between these techniques, we can reasonably expect to identify the particle nature of dark matter within the coming decade